The membrane-proximal region (MPR) of herpes simplex virus gB regulates association of the fusion loops with lipid membranes

Abstract

Glycoprotein B (gB), gD, and gH/gL constitute the fusion machinery of herpes simplex virus (HSV). Prior studies indicated that fusion occurs in a stepwise fashion whereby the gD/receptor complex activates the entire process, while gH/gL regulates the fusion reaction carried out by gB. Trimeric gB is a class III fusion protein. Its ectodomain of 773 amino acids contains a membrane-proximal region (MPR) (residues 731 to 773) and two fusion loops (FLs) per protomer. We hypothesized that the highly hydrophobic MPR interacts with the FLs, thereby masking them on virions until fusion begins. To test this hypothesis, we made a series of deletion, truncation, and point mutants of the gB MPR. Although the full-length deletion mutants were expressed in transfected cells, they were not transported to the cell surface, suggesting that removal of even small stretches of the MPR was highly detrimental to gB folding. To circumvent this limitation, we used a baculovirus expression system to generate four soluble proteins, each lacking the transmembrane region and cytoplasmic tail. All retained the FLs and decreasing portions of the MPR [gB(773t) (gB truncated at amino acid 773), gB(759t), gB(749t), and gB(739t)]. Despite the presence of the FLs, all were compromised in their ability to bind liposomes compared to the control, gB(730t), which lacks the MPR. We conclude that residues 731 to 739 are sufficient to mask the FLs, thereby preventing liposome association. Importantly, mutation of two aromatic residues (F732 and F738) to alanine restored the ability of gB(739t) to bind liposomes. Our data suggest that the MPR is important for modulating the association of gB FLs with target membranes. IMPORTANCE To successfully cause disease, a virus must infect host cells. Viral infection is a highly regulated, multistep process. For herpesviruses, genetic material transfers from the virus to the target cell through fusion of the viral and host cell lipid membranes. Here, we provide evidence that the ability of the herpes simplex virus (HSV) glycoprotein B (gB) fusion protein to interact with the host membrane is regulated by its membrane-proximal region (MPR), which serves to cover or shield its lipid-associating moieties (fusion loops). This in turn prevents the premature binding of gB with host cells and provides a level of regulation to the fusion process. These findings provide important insight into the complex regulatory steps required for successful herpesvirus infection.

FIG 1

(A) Surface representation of the gB trimer. Residues of FL1 (from all 3 monomers) are shown in pink, while residues of FL2 are shown in blue. The C termini of the solved gB monomers are highlighted (yellow). The unsolved MPRs (membrane-proximal regions) (amino acids 730 to 773) of each protomer are represented as dashed, thick tan lines emanating from the gB monomer C termini. FL1 and FL2 of one protomer are labeled in the bottom view of the trimer. (B) Schematic representation of a gB protomer, showing its structural domains (Roman numerals). FL1 and FL2 are indicated with arrows, colored as in panel A. Amino acid numbers are shown along the top. The MPR is shown in tan, and its residues are in expanded view below. Residues that were mutated in this study are highlighted in boldface type with an asterisk. Deletion mutants are designated boxes and aligned below the MPR amino acids, with a dashed line indicating deleted residues in the full-length protein. sig, signal sequence; TMR, transmembrane region. (C) Schematic representation of the soluble C-terminal (C-term) truncation mutants generated in this study. The point mutations F732A and F738A are indicated as black bars.

FIG 2

Characterization of gB MPR deletion mutants. (A) Full-length MPR mutants were expressed in mammalian cells, and total cell extracts were analyzed by denaturing Western blotting. Blots were probed with the anti-gB PAb R68. (B) Protein surface expression as detected by CELISA (black bars). Transfected CHO-K1 cells were fixed with 3% paraformaldehyde, then incubated with the anti-gB PAb R69 and goat anti-mouse antibody conjugated to horseradish peroxidase (GAM-HRP). Cells transfected with empty vector DNA were used as a negative control, and this value was subtracted from the other experimental samples. Quantitative cell-cell fusion assay (gray bars) was performed with cocultivation of target CHO-K1 cells (expressing the luciferase protein and the HSV [herpes simplex virus] receptor HVEM) with effector CHO cells (expressing T7 polymerase, gD, gH, and gL, plus either WT gB, mutant gB, or empty vector DNA). Cell extracts were tested for light production 18 h later. Percent WT was calculated as follows: for CELISA = (sample absorbance/WT absorbance) × 100; for fusion assay = (RLU of test sample/RLU of WT) × 100 where RLU stands for relative light units.

FIG 3

Soluble gB MPR-containing proteins are expressed and folded correctly. (A) Four gB mutants (739t, 749t, 759t, and 773t) were cloned and expressed in a baculovirus expression system as secreted forms with each protein truncated after the indicated amino acid. Proteins were detected with the PAb R69 and visualized by denaturing or “native” Western blots. (B) Reactivity of gB MPR-containing proteins with the conformation-dependent MAbs SS55 or DL16 via “native” Western blotting. The positions of molecular mass standards (in kilodaltons) are shown to the left of the blots in panels A and B. (C) Liposome flotation assay. Purified soluble glycoproteins were incubated with liposomes for 1 h at 37°C. Samples were adjusted to 1 M KCl, incubated for an additional 15 min, layered beneath a discontinuous 5 to 40% sucrose gradient, centrifuged for 3 h, and then fractionated. The top, liposome-containing fraction was analyzed by dot blotting with the PAb R68. lip, liposomes.

FIG 4

Liposome binding assay and biosensor analysis. (A) Liposomes are injected and allowed to flow across flow cell 1 (Fc1) and Fc2, with binding at saturation (~8,500 response units  [RU]). Soluble gB(730t) is then injected across Fc2 at 5 µl/min for 240 s, and binding to liposomes is measured as an increase in RU (response units); this portion of the graph (boxed) is shown and analyzed in subsequent figures. A double-slash denotes a break in the x axis (no injections were performed during that time period). (B) Response curve showing binding of control proteins gB(730t) (positive control), fusion loop mutant gB-W174R(730t) (negative control), and soluble gD (negative control). (C and D) Twofold serial dilutions of gB(730t) (C) or gB(773t) (D) were injected across the liposome-coated flow cell at 10 µl/min for 2 min to evaluate ligand-liposome association. After each injection, the surface preparation protocol was performed to remove protein and liposomes from the chip, regenerating the surface to the RU baseline. (E) Each soluble protein (156 nM) was allowed to flow across an L1 chip containing immobilized liposomes as described above for panel A. The flow rate was 5 µl/min. (F) Bar graph representation of gB-liposome binding via biosensor. Values are averages of at least 3 separate experiments, with samples presented as percent binding of gB(730t) (set at 100%). Error bars represent standard errors of the means.

FIG 5

Gel filtration/size exclusion chromatography of gB(730t) (A) and gB(773t) (B). Fractions containing mostly trimeric species (boxed in the Western blot) were tested via biosensor for liposome binding at 156 nM gB for 240 s. Fractions containing high-molecular-weight (HMW) species, samples to the right of the box on the Western blot) were excluded from study. “Total” refers to the gB sample before it underwent gel filtration.

FIG 6

(A) Model of a possible interaction between the gB fusion loops and MPR. The gB trimer is rendered in cartoon form in gray, focusing on the fusion loops (FL1 shown in pink and FL2 shown in blue) and C terminus of one of the protomers (H724, yellow spheres). The side chains of both fusion loops are shown. The MPR is depicted as a thick tan dashed line, with the aromatic residues F732 and F738 represented as circles. (B) Native Western blot of soluble gB proteins, probed with the indicated MAbs. The positions of molecular mass standards (in kilodaltons) are shown between the two blots. (C) Native Western blot of gel filtration fractions as described in the legend to Fig. 5. (D) gB-liposome binding assay (biosensor analysis) as described in the legend to Fig. 4A. Soluble gB(730t) served as the positive control and soluble gD as the negative control.